Response—Cretaceous Extinctions

See allHide authors and affiliations

Science  21 May 2010:
Vol. 328, Issue 5981, pp. 975-976
DOI: 10.1126/science.328.5981.975

The Letters by Archibald et al., Keller et al., and Courtillot and Fluteau question our conclusion that the Cretaceous-Paleogene mass extinction was caused by the asteroid impact at Chicxulub. All three Letters stress that Deccan flood basalt volcanism played a major role in the extinction. Keller et al. and Archibald et al. also mention that climate change was a factor, and Archibald et al. point to marine regression as well.

We disagree with the hypothesis that volcanic activity can explain the extinction. First, geographically extensive biotic records of marine microfossils and terrestrial pollen and spores that reveal the nature of the Cretaceous-Paleogene (K-Pg) mass extinction with the greatest fidelity do not contain evidence of accelerated extinction rates during the last 400 thousand years of the Cretaceous [our Review and (1, 2)] and therefore do not support the idea that the biosphere was somehow destabilized by Deccan volcanism. In fact, plant macrofossils record a diversification during this time (2). Studies of the last 1.5 million years of the Cretaceous from North America, Europe, and Asia [e.g., (3, 4)] are compatible with a sudden extinction scenario for non-avian dinosaurs. Moreover, the constancy of late Maastrichtian open ocean sedimentation (as indicated by climate cycles driven by regular oscillations in Earth's orbit) does not provide evidence for overall declining productivity or instability in marine ecosystems preceding the boundary [e.g., (5)].

Second, recent studies suggest that the emplacement of the Deccan flood basalts took place during multiple (∼30) large eruptive pulses, most of which predate the K-Pg boundary by several hundred thousand years (6). In contrast, others have argued that “activity in the continental flood basalt province as a whole is likely to have been quasi-continuous” (7). Nevertheless, it is extremely difficult to reconcile the protracted Deccan flood basalt eruption history with a single abrupt mass extinction horizon exactly at the K-Pg boundary. Although it is well documented that the Chicxulub impact event coincided precisely with sudden paleontological and paleoenvironmental changes and the K-Pg mass extinction [our Review and (8, 9)], there are no comparable data demonstrating that a major pulse of Deccan volcanism coincided with the mass extinction. Moreover, it remains to be explained why one eruptive event would have resulted in mass extinction, whereas multiple earlier eruptive events of comparable magnitude and duration occurring up to 500 thousand years before the K-Pg boundary (6) left few global environmental traces [e.g., (1, 2)].

Third, rates of sulfur injections are critically important to discriminating between environmental consequences of impact versus those of volcanism because the residence time of sulfur in the atmosphere is short (10). Courtillot and Fluteau claim that we misrepresent their 2009 paper (6). However, the paper includes exactly the numbers (reported as “0.1 to 1 Gt/a sulfur dioxide”) we stated. We did not note their finding that the sulfur was released “over durations possibly as short as 100 years for each single eruptive event” (6) because this does not affect our conclusions. Maintaining such a sulfur release for 100 years would indeed result in a total sulfur release of 50 Gt, which is in the order of the lowest estimate for Chicxulub impact (see our Review). However, sulfur is removed from the atmosphere continuously (10) and therefore any accumulation in the atmosphere is unsupported, contrary to the claim made by Courtillot and Fluteau. We also emphasize that the instantaneous release of 100 to 500 Gt sulfur is only one consequence of the Chicxulub impact, and the K-Pg boundary mass extinction is likely the result of a combination of several impact-induced environmental effects (including the release of sulfur, soot, dust, and other effects, as noted in our Review), whereas the Deccan flood basalt hypothesis relies exclusively on the injection of sulfur dioxide (6).

The Chicxulub Crater.

A computer-generated gravity map image shows the Chicxulub Crater on Mexico's Yucatan Peninsula.


With regard to Archibald et al.'s and Courtillot and Fluteau's comments about other Phanerozoic mass extinction events that co-occurred with the emplacement of flood basalt provinces, we note that these extinction events are commonly associated with oceanic anoxia, calcification crises, and strong global warming—none of which is observed at the K-Pg boundary (2, 1013). Furthermore, there is an absence of mass extinctions during several large flood basalt eruptions (10, 14). Each mass extinction event should be considered relative to the record for that event [e.g., (12)], and we stress the unique aspects of the K-Pg boundary record. Chicxulub is by far the largest known impact event in the Phanerozoic, and the projectile hit an extraordinarily thick sulfur-rich sedimentary sequence (see our Review). The absence of evidence for impact phenomena at other mass extinctions, discussed by Archibald et al., is irrelevant for our synthesis of the stratigraphy and biotic response to the specific Chicxulub impact event.

Our work in no way diminishes the importance of gaining a better understanding of the environmental consequences of massive volcanism. We do not doubt that such volcanism can significantly perturb the global environment. However, a robust correlation between mass extinction and flood basalt volcanism as suggested by Courtillot and Fluteau is unlikely [see reviews of (10, 14)].

Keller et al. and Archibald et al. mention that climate change contributed to the extinction. As outlined in our Review and in (1, 2), climate fluctuations during the latest Maastrichtian (minor warming and subsequent cooling) and the associated faunal and floral consequences are clearly separated from the abrupt mass extinction event at the K-Pg boundary.

In response to Archibald et al.'s point about marine regressions, we note that marine mass extinctions may have coincided with global sea-level changes [e.g., (15)]. However, because sea-level changes are numerous (15), this association seems coincidental rather than causal (16). Sea-level change also fails to explain the disruption of vegetation and the faunal change observed in terrestrial environments at the K-Pg boundary (1).

We disagree with the comments of Keller et al. regarding the association between Chicxulub impact ejecta and the K-Pg boundary, and we point out that our Review addressed all of the issues to which they refer. Our Review integrated new data with previous work in the peer-reviewed literature to provide substantial corroborating evidence for a global correlation of the Chicxulub impact with the K-Pg boundary.


Stay Connected to Science

Navigate This Article